Knowledge-based information fusion for improved situational awareness

This project aims to investigate knowledge-based approaches to the problem of information fusion. We show how domain-specific knowledge, in conjunction with semantic Web services, can be used to improve situational awareness in the operational deployment of humanitarian operations, especially when such operations occur against a contextual backdrop of ongoing military conflict. Information, harvested from a variety of physically disparate and semantically heterogeneous data sources, is used to establish a common repository of semantically rich, conceptual-level representations of real-world events. Reasoning services, deployed over the knowledge repository, are used to intelligently fuse information, taking into account factors such as the level of trust and confidence assigned to specific information sources. Selective attention to information of relevance to particular users helps avoid situations of information overload that might otherwise result from the information fusion process. We believe this approach to have general applicability in a wide variety of military contexts.